Comparative efficacy and complications of vena caval filters.

Purpose: A variety of vena caval filters (VCFs) are available for usage. The choice of filter type depends on physician preference and certain patient variables. An evaluation of the different VCFs used in our institution was done to compare their efficacy and complication rates.

Methods: The medical records of all patients who underwent insertion of a VCF from January 1987 to June 1993 at the University of Iowa Hospitals & Clinics and the affiliated Veterans Administration Medical Center were reviewed. One hundred ninety-nine VCFs were placed in 196 patients (123 males, 73 females), with a mean age of 61 years (range 13 to 87 years). Thirty-five (18%) VCFs (30 stainless steel Greenfield filters [SGFs] and five titanium Greenfield filters with modified hook [TGF-MHs]) were inserted in the operating room via an open technique. The remaining 164 VCFs (82%) were inserted in the radiology suite by a percutaneous technique (38 SGF, 23 TGF-MH, 51 Vena Tech filters [VTFs], 48 Bird's nest filters [BNFs] and 4 Simon Nitinol filters). Thromboembolic risk factors in these 196 patients included malignancy (99), trauma (21), recent surgery (27), cerebrovascular accident with paralysis (6), and miscellaneous conditions (43). Indications for VCF placement included a contraindication to anticoagulation (92), complication of anticoagulation (44), failure of anticoagulation (26), prophylaxis (31), adjunct to pulmonary embolectomy (1), noncompliance (1), hemodynamically unstable patient (1), and prior VCF complication (3). Mean follow-up of the patients was 12 months (range 0 to 87 months). Because there were only four Simon Nitinol filters inserted during the study period, they were excluded from further analysis.

Results: A comparative analysis revealed that there was a significantly higher incidence of symptomatic IVC thrombosis with the use of the BNF (n = 7) (14.6%) versus the SGF (n = 0) (0%), TGF-MH (n = 1) (3.6%), or VTF (n = 2) (4%) (p < 0.05 by chi-squared testing). The VCF-related mortality rate was also higher with the BNF (n = 5) (10.9%) versus the SGF (n = 1) (1.5%), TGF-MH (n = 1) (3.6%), or VTF (n = 0) (0%) (p < 0.05 by chi-squared testing). However there was no significant difference in the occurrence of clinically apparent recurrent pulmonary embolism during follow-up between the four different filter types (2 [4.2%] BNF, 3 [4.4%] SGF, 1 [3.6%] TGF-MH, and 1 [2%] VTF).

Conclusion: These data indicate that the use of the BNF was associated with increased morbidity and mortality rates compared with the use of the SGF, TGF-MH, and VTF filters.